FCM Combined with CNN for Evaluating the Metering Performance of Electric Energy Meters

To ensure the stable performance of smart meters under different measurement environments, an novel method for evaluating the metering performance of electric energy meter based on big data analysis is proposed. By statistically analyzing measurement errors under standard laboratory conditions and combining this with field meter operation data and environmental factors, error corrections are made. First, the Fuzzy C-Means(FCM) clustering method is used to perform cluster analysis on both laboratory and field operation environments. Then, convolutional neural network (CNN) is employed to analyze and train voltage, current, power factor, and environmental data, establishing a mapping relationship between meter measurement errors and these factors. Finally, error correction is carried out on the obtained field smart meter data. The results show that the proposed method can correct the measurement errors of smart meters, verifying the feasibility of the proposed method.